Denoising Autoencoding System

A Denoising Autoencoding System is a feedforward network training system that implements a denoising autoencoding algorithm to solve a ...

• Context:
  • It can be used to create a Stacked Denoising Autoencoding System.
  • …
• Example(s):
  • TensorFlow's run_autoencoder.py.
  • …
• Counter-Example(s):
  • Deep Autoencoder.
• See: Learned Neural Network.

---

---

References

2016

• http://deep-learning-tensorflow.readthedocs.io/en/latest/#denoising-autoencoder
  • QUOTE: Cmd example usage:

    python command_line/run_autoencoder.py --n_components 1024 --batch_size 64 --num_epochs 20 --verbose 1 --learning_rate 0.05 --corr_type masking --corr_frac 0.5 --enc_act_func sigmoid --dec_act_func sigmoid --loss_func cross_entropy --opt momentum --momentum 0.9 --main_dir dae-models --model_name dae1024

    This command trains a Denoising Autoencoder on MNIST with 1024 hidden units, sigmoid activation function for the encoder and the decoder, and 50% masking noise. You can also initialize an Autoencoder to an already trained model by passing the parameters to its build_model() method.

---